ii8 



SCIENCE. 



[Vol. XIX. No. 473 



SCIENCE; 



centage of this mean to the actual distance, 

 are in decimals of an inch. 



All distances 



A WEEKLY NEWSPAPER OF ALL THE ARTS AND SCIENCES. 



PUBLISHED BY 



N. D. C. HODGES. 



874 Broadway, New York. 



Subscriptions.— United States and Canada $3.50 a year. 



Great Britain and Europe 4.50 a year. 



Communications wiil be welcomed from any quarter. Abstracts of scientific 

 papers are solicited, and one hundred copies of the issue containing such will 

 be mailed the author on request in advance. Rejected manuscripts will be 

 returned to the authors only when the requisite amount of postage accom- 

 panies the manuscript. Whatever is intended for insertion must be authenti- 

 cated by the name and address of the writer; not necessarily for publication, 

 but as a guaranty of good faith. We do not hold ourselves responsible for 

 any view or opinions expressed in the communications of our correspondents. 



Attention is called to the "Wants" column. All are invited to use it in 

 soliciting information or seeking new positions. The name and address of 

 applicants should be given in full, so that answers will go direct to them. The 

 " Exchange " column is likewise open. 



For Advertising Rates apply to Henry F. Taylor, 47 Lafayette Place, New 

 Tork. 



ESTIMATES OF DISTANCE. 



Herbert Nichols, la his experiments on "The Psychol- 

 ogy of time" {American Journal of Psychology, April, 

 1891), has shown that estimates of time intervals are influ- 

 enced by immediately preceding estimates, so that, in general, 

 intervals are judged to be longer after practice on estimating 

 an actually longer interval than when no such practice pre- 

 cedes, and shorter after practice on a shorter interval. The 

 experiments about to be described were undertaken to see 

 whether the same rule applies to estimates of distance. They 

 show no such effect, perhaps because the intervening prac- 

 tice was not sufficiently sustained to affect the judgment. 

 But the results are interesting for several reasons, and they 

 are therefore given below. 



The mode of experimenting was as follows: On each of 

 three sheets of unruled paper (about six by nine inches) was 

 placed a pair of pencil dots; on the first these were 4.03 

 inches apart; on the second .92 of an inch; on the third ex- 

 actly the same distance as on the first. Without being told 

 the object of the experiments, the person to be experimented 

 on was shown the first pair of dots, allowed to look at them 

 as long as he pleased, and then, the paper being taken away, 

 told to make from memory, on a slip 9X J inches, two dots at 

 the same distance apart, as nearly as he could. This was 

 repeated on a fresh sheet, without his looking at the model 

 again, and so on till he had made ten trials. The same thing 

 was then repeated with the second and third sheets. 



The following table shows the results, the first column 

 giving the difference between the actual distance of the dots 

 and the average of the ten estimates in each series; the second 

 column the percentage of this difference to the actual dis- 

 tance; the third the mean deviation of the estimates from the 

 average (taken always as positive); and the fourth the per- 



The degree to which the absolute value of the errors de- 

 pends on previous training is plainly shown; for instance, 

 L. F., in whose case they are remarkably small, is the 

 daughter of a well-known artist and herself accomplished in 

 the use of the pencil, while A. L. B. is a boy five years of 

 age. The consistency of the estimates seems, however, to 

 depend much less on training, as shown in the third column, 

 the ratio of A. L. B's. mean deviations to those of L. P. 

 being about 1.,^, 1.7, and 6.8 for the three series respectively, 

 while the ratios of their errors (from the first column) are 

 27, 4, and 131. In the cases of S. S., A. L. B., and L. B. 

 the errors are nearly proportional to the actual length of the 

 intervals, which would seem the natural rule; but in the other 

 cases there seems a tendency toward making errors of the 

 same absolute value in estimating both short and long inter- 

 vals. A. L. B. , whose absolute errors are far the largest, 

 keeps them most nearly proportional. The mean deviations 

 are much more generally proportional to the intervals, the 

 most noticeable exception being that of J. S. — also the chief 

 exception to proportionality in the former case. 



Arthur E. Bostwick. 



THE LATEST ADVANCES IN SPECTRUM PHOTOG- 

 EAPIIY. 



A LETTER just received by the present writer from Mr. 

 Victor Schumann of Leipzig, whose work in the domain of 

 spectrography is less widely known and appreciated than it 

 deserves to be, reveals such surprising advances within the 

 past year in photographing radiations in the ultra-violet 

 spectrum, that I am impelled to present the following sum- 

 mary of Mr. Schumann's results. 



More than two years ago he demonstrated the remarkable 

 absorptive effect of air upon very short vibrations, so gi'eat, 

 indeed, that even the air within the tubes of the spectrograph 

 was a serious obstacle to the investigation. However, he 

 was able, with the apparatus then at hand, to demonstrate 

 the existence of lines up to and beyond wave-length 1,852 by 

 photography, using the light of the aluminum spark. 



With the fine skill and ingenuity which has ever charac- 

 terized his work, Mr. Schumann has since constructed a 

 spectrograph exhausted of air, with lenses and prism of 

 white fluor-spar. The source of light for these researches 

 was the hydrogen Geissler tube. With the "exhausted 

 spectroscope," as it is termed, and plates of proper sensitive- 

 ness, Mr. Schumann finds the photographic action of the 

 spectrum beyond wave-length 1,852 very strong indeed. It 

 is composed of fourteen groups of lines, including altogether 

 about six hundred lines. The boundary of this hitherto en- 



